distillbert ktrain‘太多的值无法解包’

Question

我正尝试在Colab中用ktrain运行DistilBert，但是我得到了“错误，太多的值无法解包”。我正在尝试执行有毒评论分类，我从CivilComments上传了'train.csv‘，我可以运行BERT，但不能运行DistilBert

#prerequisites:
!pip install ktrain
import ktrain
from ktrain import text as txt
DATA_PATH = '/content/train.csv'
NUM_WORDS = 50000 
MAXLEN = 150 
label_columns = ["toxic", "severe_toxic", "obscene", 
                 "threat", "insult", "identity_hate"]

如果我只是使用'bert‘进行预处理，它可以工作得很好，但是我不能使用distilbert模型。在使用distilbert进行预处理时，我得到以下错误：

 (x_test, y_test), preproc = txt.texts_from_csv(DATA_PATH, 'comment_text', label_columns=label_columns, val_filepath=None, max_features=NUM_WORDS, maxlen=MAXLEN,  preprocess_mode='distilbert')

“解包的值太多了，应该是2”，如果我用bert替换distilbert，它可以正常工作(代码如下)，但我不得不使用bert作为模型，使用bert进行预处理工作正常：

(x_train, y_train), (x_test, y_test), preproc = txt.texts_from_csv(DATA_PATH, 'comment_text', label_columns=label_columns, val_filepath=None, max_features=NUM_WORDS, maxlen=MAXLEN,  preprocess_mode='bert')

这个没有错误，但我不能使用distilbert，如下所示：

示例：model = txt.text_classifier('distilbert', train_data=(x_train, y_train), preproc=preproc)错误消息：if 'bert' is selected model, then preprocess_mode='bert' should be used and vice versa

我想使用(x_test, y_test), preproc = txt.texts_from_csv(DATA_PATH, 'comment_text', label_columns=label_columns, val_filepath=None, max_features=NUM_WORDS, maxlen=MAXLEN, preprocess_mode='distilbert')和distillbert模型，如何避免错误‘值太多，无法解包’

代码所基于的链接: Arun Maiya (2019)。ktrain: Keras的轻量级包装器，用于帮助训练神经网络。https://towardsdatascience.com/ktrain-a-lightweight-wrapper-for-keras-to-help-train-neural-networks-82851ba889c。

Answer 1

如this example notebook所示，当将texts_from_*指定为模型时，TransformerDataset函数将返回NumpyArrays对象(而不是preprocess_mode='distilbert' )。所以，你需要这样做：

trn, val, preproc = txt.texts_from_csv(DATA_PATH, 'comment_text', label_columns=label_columns, val_filepath=None, max_features=NUM_WORDS, maxlen=MAXLEN,  preprocess_mode='distilbert')